Re: [v6ops] [dhcwg] SLAAC renum: Problem Statement & Operational workarounds

[Ole Troan <[e-mail]>]

>> I wonder if introducing suggestions (1) or (2) aren't pushing existing
>> clients to expose new bugs they may have? Suggestion (4) seems
>> reasonable to me.
> 
> Clients that have bugs should be updated.  Routers that aren’t updatable should be binned.   Sites that have massive numbers of non-updatable routers that can’t be updated don’t have to turn the feature on.   I think it’s fine to make this a server configuration parameter.
> 
> The benefit to doing this correctly is that it’s arguably already correct.  There are likely routers out there that will already do the right thing when they get a deprecated prefix.   The most likely use case for a deprecated prefix is not that it has a valid lifetime of zero, but that it has a preferred lifetime of zero: a graceful renumbering event.   DHCPv6 should already support graceful renumbering.   This was a design goal.

Right, why wouldn't this just be (accellerated) graceful renumbering as already described in 3633/8415.
I don't understand why this would expose bugs or require a new option?

Cheers,
Ole